The original description of the correlation between birthweights of mothers and their infants was followed by the description of the association between large maternal birth weight and delivery of a macrosomic (>4000 gram) infant. Study of other fetal growth parameters, including length and head circumference, demonstrated that infants of low birth weight mothers were both shorter and lighter than infants of larger mothers, but that the infants were normally proportioned. In ensuing studies, birth certificates of infants born in Tennessee between 1979 to 1984 were matched with those of their mothers, who were born in Tennessee between 1959 to 1966. Maternal and infant birth weights were again shown to be correlated. In addition, women who were themselves of low birth weight were up to 4 times as likely to have a small-for-gestational age infant as were women who weighed 4000-4499 grams at birth, but low birth weight women were less than twice as likely to have a preterm infant. A group of Swedish women, born from 1955 to 1965, was next studied. Women who themselves were small-for gestational age at birth were at increased risk of giving birth to both small-for-gestational age and preterm infants. Women who were preterm at birth were not at increased risk of either outcome. Among women born in the Danish Perinatal Study (1959-61), women born SGA were found to be at doubled risk of having an SGA child, but women born preterm were not at increased risk of giving birth to a preterm child. Analysis of data from girls who were born in the 1960's as subjects in the Collaborative Perinatal Project is currently underway. Small-for-gestational age, preterm and control girls have been located and interviewed. Hospital records of their deliveries have also been retrieved.